A Semantic Web Approach to Low-Level Features in Images

• Miguel Bento Alves Instituto Politécnico de Viana do Castelo, Portugal
• Carlos Viegas Damásio Univ. Nova de Lisboa, Portugal
• Nuno Correia Univ. Nova de Lisboa, Portugal

Resumo

We present a Semantic Web based approach that meets the requirements for images retrieval based on low-level features. For that, we developed a prototypical implementation using a Semantic Web framework, linking open data ontologies and image processing libraries. We created an ontology to support the low-level features data of images. Furthermore, we created semantic rules to define concepts based on lowlevel features. We show how the different mechanisms of the Semantic Web may help multimedia management, both for storage and in retrieval tasks.